Ischemic stroke continues to be a major public health problem exacting billions of dollars from society in medical care costs and lost wages, as well as incalculable suffering to victims and their families. Recently, thrombolytic therapy with intravenous recombinant tissue plasminogen activator (rt-PA) was shown to improve neurological and functional outcome in patients with acute ischemic stroke who could be treated with 3 hours of onset. However, the risk of symptomatic intracranial hemorrhage complicating rt-PA treatment was 6.4%, and 39% of patients either died or were severely disabled despite treatment. A novel, second generation thrombolytic drug, TNK-TPA, has been specificity engineered to be more fibrin specific than rt-PA, and has been shown in experimental models of thrombosis to be more potent and active more quickly than rt- PA. In experimental stroke, cerebrovascular clots were lysed more quickly, and the incidence of intracranial hemorrhage was reduced. In human studies of myocardial infarction, the drug is well tolerated, and the incidence of intracranial hemorrhage complicating treatment is not greater, and may be less than that with rt-PA. This proposal describes a multi-center, open-label, pilot dose-escalation safety trial of TNK-TPA in patients with acute ischemic stroke who can be treated within 3 hours. The primary hypothesis to be tested is that TNK-TPA may be safely administered to patients with acute ischemic stroke at doses which may be associated with improvement in neurological outcome. The primary endpoint of the study will be the incidence of symptomatic intracranial hemorrhage within 36 hours of treatment. Safety monitoring and sample size calculations are designed so that enrollment will cease if the observed intracranial hemorrhage rate exceeds that reported with rt-PA treatment of stroke. Drug activity will be gauged by the proportion of patients with major early neurological improvement as determined by an 8 or more point improvement on the National Institutes of Health Stroke Scale or a perfect score of 0 at 24 hours after stroke onset. Outcome at 3 months will be assessed with the Barthel Index, the modified Rankin Sale, and a global outcome statistic to facilitate sample size calculations for subsequent studies. Sixty to 100 patients will be enrolled in dosage tiers of up to 25 patients each.